DE102013106895A1 - Light microscopic method for the localization of point objects - Google Patents

Abstract

Described is a light microscopic method for localization of point objects (26) in a sample (14), in which the sample (14) by means of imaging optics (12) is imaged onto a detector (16); and locate the point objects (26) contained within the sample (14) within the depth-of-field (18) by generating lateral x / y based on a sample image produced by imaging the sample (14) on the detector (16) Positions of the point objects (26) in the direction perpendicular to the optical axis (O) are determined, wherein the depth of field (18) in the object space relative to the sample (14) along the optical axis (O) at least once by a predetermined axial z-displacement (Δz) shifted and with axially shifted depth of field (18), the sample (14) again imaged on the detector (16) and another sample image is generated; the lateral x / y positions of the point objects (26) are again determined on the basis of the further sample image; determining lateral x / y position deviations between the lateral x / y positions of the respective same point objects (26) determined on the basis of the different sample images; and in dependence on the determined lateral x / y position deviations a correction information is generated, by means of which the lateral x / y positions of the point objects (26) are corrected.

Description

The invention relates to a light microscopic method for localizing point objects in a sample, in which the sample arranged in an object space is imaged onto a detector by means of imaging optics having a depth of field of predetermined axial z-dimension along their optical axis in the object space the point objects contained in the sample are located within the depth of field by determining lateral x / y positions of the point objects in the direction perpendicular to the optical axis based on a sample image produced by imaging the sample on the detector.

More recently, light microscopic imaging techniques have been developed that can be used to represent sample structures that are smaller than the diffraction limit of classical light microscopes based on a sequential, stochastic localization of individual markers, especially fluorescence molecules. Such methods are described, for example, in WO 2006/127692 A2 ; DE 10 2006 021 317 B3 ; WO 2007/128434 A1 . US 2009/0134342 A1 ; DE 10 2008 024 568 A1 ; WO 2008/091296 A2 ; "Stochastic optical reconstruction microscopy (STORM) sub-diffraction-limiting imaging", Nature Methods 3, 793-796 (2006), MJ Rust, M. Bates, X. Zhuang ; "Resolution of lambda / 10 in fluorescence microscopy using fast single molecule photo-switching", Geisler C. et al, Appl. Phys. A, 88, 223-226 (2007) , This new branch of microscopy is also called localization microscopy. The methods used are known in the literature, for example under the designations (F) PALM ((Fluorescence) Photoactivation Localization Microscopy), PALMIRA (PALM with Independently Running Acquisition), GSD (IM) (Ground State Depletion of Individual Molecule Return) Microscopy) or (F ) STORM ((Fluorescence) Stochastic Optical Reconstruction Microscopy).

The new method has in common that the sample structures to be imaged are prepared with point objects, so-called markers, which have two distinguishable states, namely a "bright" state and a "dark" state. For example, when fluorescent dyes are used as markers, the bright state is a fluorescent state and the dark state is a non-fluorescent state.

In preferred embodiments, such as in the WO 2008/091296 A2 and the WO 2006/127692 A2 , photo-switchable or photoactivatable fluorescent molecules. Alternatively, as in the DE 10 2006 021 317 B3 , inherent dark states can be exploited by standard fluorescent molecules.

To image sample structures with a resolution that is higher than the classical resolution limit of the imaging optics, a small subset of the markers is now repeatedly transferred to the bright state. In the simplest case, the density of the markers forming this active subset is to be selected so that the mean distance between adjacent markers in the bright and thus light-microscopically imageable state is greater than the resolution limit of the imaging optics. The markers forming the active subset are formed on a spatially resolving light detector, e.g. a CCD camera, so that a light distribution in the form of a light spot is detected by each punctiform marker whose size is determined by the resolution limit of the optics.

In this way, a plurality of raw data frames are recorded, in each of which a different active subset is mapped. In an image evaluation process, the center of gravity positions of the light distributions are then determined in each raw data single image, which represent the punctiform markers located in the bright state. The center of gravity positions of the light distributions determined from the raw data individual images are then compiled in an overall representation in the form of an overall image data set. The high-resolution overall image resulting from this overall representation reflects the distribution of the markers.

For a representative reproduction of the sample structure to be imaged, a sufficient number of marker signals must be detected. However, since the number of evaluable markers in the respective active subset is limited, a large number of raw data individual images must be recorded in order to completely map the sample structure. Typically, the number of raw data frames is in the range of several tens of thousands, and this range varies widely, because for complex structures far more images must be taken than for simpler structures to resolve the structures.

In addition to the above-described lateral position determination of the markers in the object plane (also referred to below as the x-y plane), it is also possible to determine the position in the axial direction (also referred to below as the z-direction). With axial direction is meant the direction along the optical axis of the imaging optics, ie the main propagation direction of the light.

Three-dimensional localizations are from so-called "Particle tracking" experiments known as in kajo et. al., 1994, Biophysical Journal, 67, Holtzer et. al., 2007, Applied Physics Letters, 90 and Toprak et al., 2007, Nano Letters, 7 (7). They have also been used in imaging techniques based on the switching and localization of single molecules described above. Reference is made to Huang et al, 2008, Science, 319 and Juette et al., 2008, Nature Methods. The prior art is further disclosed in Pavani et al., 2009, PNAS, 106 , referenced.

A localization of a point-shaped object in the z-direction can in principle be carried out by evaluating the change in a detected on the detection surface of the camera light spot, which is visible when the point object moves out of the detection surface to the optically conjugate focus or focal plane. In the following, a point object is to be understood as an object whose dimensions are smaller than the diffraction-related resolution limit of the imaging optics, in particular of the detection objective. In this case, the detection lens images such an object into the image space in the form of a three-dimensional focus light distribution. The focus light distribution generates on the detection surface of the camera, a light spot, which is characterized by the so-called "point spread function", ie point mapping function or short PSF. Now, if the point object in the z-direction is focused by the focus, i. moved perpendicular to the focus plane, so change the size and shape of the PSF. By analyzing the detection signal corresponding to the detected light spot with respect to the size and shape of the PSF, it is possible to obtain conclusions about the actual z position of the object.

If the point object is located too far away from the focus plane, the light spot produced on the detection surface of the camera is so blurred that the corresponding measurement signal can no longer be perceived within the usual measurement noise. Thus, in the object space in the z direction, there is an area around the central focal or focus plane within which a point object on the detection surface generates a light spot which is still sharp enough to be able to be evaluated for locating the point object in the z direction , This area containing the sharpness plane in the z direction will be referred to as "depth of field" hereinafter.

In the case of a three-dimensional localization, however, there is the fundamental problem that the PSF originating from a point object is symmetrical with respect to the detection surface. This means that the PSF changes when the point object is moved out of the focus plane so that the distance of the point object to the focus plane can be determined. However, the change of the PSF is symmetrical to both sides of the focus plane, so that it can not be decided on which side of the focus plane the point object is within the depth of field.

There are various methods known how to deal with the above-described problem. Examples are methods known in the art as "Astigmatism Methods" (the above-referenced Kajo et al., Holtzer et al., And Huang et al.), "Bi-Plane" methods (see Toprak et al., And Juette et al.), And "Double helical method" (see Pavani et al.) be designated. This method has in common that for localizing the point object in the z direction of the light spot generated on a detector is analyzed to determine a characteristic and this characteristic is assigned a z-position of the point object. This assignment takes place on the basis of an assignment information determined in advance, which relates the parameter to the z position of the point object. As a parameter, for example, as in the astigmatism method, a size which characterizes the shape of the light spot or, as in the case of the bi-plane method, a size which relates the extents of two light spots to each other, comes from one and the same Resulting light spot and are generated on detection surfaces whose associated focus planes are offset in the object space in the z-direction to each other.

In localization microscopy, where resolutions of far below 100 nm, and sometimes even down to a few nm are achieved, optical aberrations, which inevitably occur in any imaging optics, are a significant problem. During classical diffraction-limited microscopy, in the resolutions measured in the object space approximately in the range of 250 nm can be achieved, the aberrations can be sufficiently minimized by exact lens production or additional correction elements, this is not readily possible in the localization microscopy. There, the resolution is so high that the remaining aberrations are of considerable relevance. Examples of such aberrations are chromatic aberrations, spherical aberrations or lateral field distortions, i. Aberrations that lead to distortion of the PSF in a plane perpendicular to the optical axis. An example of lateral field distortion is coma.

The object of the invention is to develop a light microscopic method for the localization of point objects of the type mentioned so that lateral field distortions are reliably corrected with the least possible technical effort.

The invention achieves this object according to claim 1, characterized in that the depth of field within which the point objects are located in the object space relative to the sample along the optical axis at least once by a predetermined axial z-displacement, which is smaller than the axial extent of the depth of field , shifted and with axially shifted depth of field, the sample imaged by the imaging optics again on the detector and another sample image is generated; on the basis of this further sample image, the lateral x / y positions of the point objects are again determined; determining lateral x / y positional deviations between the lateral x / y positions of the respective same point objects determined on the basis of the different sample images; and correction information is generated in dependence on the determined lateral x / y position deviations, on the basis of which the lateral x / y positions of the point objects, which have been determined on the basis of at least one of the different sample images, are corrected.

In contrast to conventional solutions, in which attempts are made to correct the aberrations solely by means of apparatus, in particular by the most error-free production of the optical elements used in the imaging optics, the invention selects a way in which the correction of aberrations from the analysis the localized positions themselves. Also, no additional calibration or a measurement intended specifically for the determination of optical errors is required, as a result of which the technical complexity is considerably reduced.

The solution according to the invention provides for a shift of the depth of field of focus along the optical axis of the imaging optics such that there is a certain overlap along the optical axis between the original depth of field and the shifted depth of field in the object space. This overlap is achieved by making the axial z-displacement, by which the depth of field is moved along the optical axis, smaller than the axial z-dimension of the depth of field. The z-adjustment path is, for example, in the range of 5 to 90%, 10 to 80%, 15 to 70%, 20 to 60% or 25 to 50% of the axial z-dimension of the depth of field. It goes without saying that these ranges of values are only to be understood as examples.

The displacement of the depth of field according to the invention about the axial z-displacement, which is smaller than the axial extent of the depth of field, is thus to be understood that the two depth of field considered, namely the original and the shifted depth of field, along the optical axis, the above-mentioned overlap exhibit. This means that the invention also covers a sequence of shifts in the depth of field in which, in a single step, the depth of field is shifted by an adjustment path that is greater than the extension of the depth of field, provided that the sequence of steps as a whole results in that between the considered Depth of field areas the ultimately desired axial overlap is realized.

The invention proposes to record not only a single sample image of a given three-dimensional structure of point objects, but at least one further sample image with a depth of field displaced along the optical axis so that one and the same point objects are imaged in different z-positions of the depth of field.

Due to the lateral field distortion to be corrected according to the invention, the point objects mentioned in the two sample images have position deviations in their lateral x / y positions as a function of their respective z-position. From these lateral x / y position deviations, correction information can now be generated which is a measure of the lateral field distortion. This correction information can then be used to correct the aberrations of the lateral x / y positions of the point objects.

As already mentioned, the depth of field according to the invention in the object space means an area in the z-direction about the central focal or focal plane within which a point object on the detector generates a light spot which is still sharp enough to localize the object Point object can be evaluated. It is not necessary to fully exploit this maximum possible depth of field. Thus, it may be useful, depending on the desired localization accuracy to deliberately reduce the depth of field and thus to exclude quite blurred, but still be evaluated light spots from the evaluation. To determine the lateral x / y positions, the light spots generated on the detector need not necessarily be strictly separated from each other spatially. By means of suitable algorithmic methods, such as are known from the literature as multi-fit methods or maximum-likelihood methods, it is also possible to analyze overlapping light distributions in such a way that the positions of the point objects can be determined.

The inventive method is not only suitable for correcting aberrations caused by the optical elements of the Imaging optics are caused. Rather, the invention also makes it possible to correct for optical disturbances caused by the sample itself. Such optical perturbations are often dependent on the location where they occur in the sample or on the ambient temperature. They are thus difficult to control with conventional methods in which, for example, optical correction elements are appropriately controlled with considerable technical outlay. This is all the more so since these conventional methods usually provide an iterative correction of aberrations in which the sample is recorded repeatedly and must therefore not change. In contrast, in localization microscopy, the signal detected by the detector constantly changes due to the flashing individual molecules, so that as a rule no sample image is exactly the same as the other one.

The lateral x / y position of a point object is to be understood below as meaning a position in the direction of the x axis and / or approximately with reference to a Cartesian coordinate system in which the z axis is parallel to the optical axis of the imaging optics is measured in the direction of the y-axis, wherein the x-axis and the y-axis are arranged perpendicular to the y-axis.

Preferably, at least one reference plane perpendicular to the optical axis is defined within the depth of field of the field, which remains stationary relative to the depth of field when the depth of field is moved. One of the sample images is defined as a reference image, based on this reference image, a comparison structure is defined, which represents at least one of those point objects that are arranged in the reference plane of the depth of focus when recording the reference image. The comparison structure is then identified in the at least one other sample image. Based on the different sample images, the lateral x / y position of the comparison structure is determined in each case. Subsequently, the lateral x / y positional deviation between the lateral x / y positions of the reference structure determined on the basis of the different sample images is determined. Finally, the correction information is generated as a function of the lateral x / y position deviation determined for the comparison structure.

The reference plane is, for example, the central focal or focus plane within the depth of field. The latter is moved along the optical axis as the depth of field moves, thus scanning the object space as it were. On the basis of one of the sample images, which is defined as a reference image, a comparison structure is defined, which is also present in the other sample images, but there in different z-positions and thus - due to the lateral field distortion - in other x / y positions within of the associated depth of field. With the aid of the comparison structure, the lateral x / y position deviations can thus be reliably determined as a function of the z position.

Preferably, the depth of field is axially displaced in several steps. In each of these steps, the lateral x / y positional deviation of the comparison structure determined on the basis of the associated sample image is determined relative to the lateral x / y position of the comparison structure determined on the basis of the reference image. As correction information, an assignment function is generated, the function values of which in each case indicate the lateral x / y position deviation of the associated comparison structure determined in the respective step as a function of its axial z position along the optical axis. The assignment function represents a correction rule which, depending on the distance of a point object from the reference plane, provides for a correction of the lateral x / y position, which exactly compensates for the mislocalization of this point object caused by the lateral field distortion.

Values of the assignment function that lie between the function values determined by the stepwise shift of the depth of field can be determined, for example, by interpolation. In this way, a continuous assignment function can be determined, although only a few discreetly determined function values are present due to the stepwise shifting of the depth of field.

In one possible embodiment, the lateral x / y positions of the point objects are corrected by image processing directly in the associated sample image. This has the advantage that the corrected positions of the point objects are immediately displayed to the user without having to go through the correction of previously stored positions and a sample image newly generated on the basis of the corrected positions.

Preferably, the comparison structure is identified in the at least one other sample image as a function of the image brightness detected on the detector, ie taking into account the total number of light spots generated on the detector that contribute to this structure. This embodiment is particularly advantageous if the x / y positions determined with the depth of field shifted are not only used for correcting positions that are previously subject to aberrations, but at the same time are used to generate a high-resolution overall localization image. This can be annoying Brightness differences in the overall localization image can be avoided.

The sum of the individual axial z-displacement paths is substantially equal to the axile z-extension of the depth of field. However, it is also possible to adjust the depth of field in the z-direction as a whole by a distance which is greater than the extent of the depth of field. If, in this case, the sample images generated in individual steps are combined to form an overall localization image after the positions of the point objects have been corrected, this covers an area in the z direction which is greater than the depth of field. Thus, in particular, complex three-dimensional structures can be imaged in high-resolution.

Preferably, the axial z-displacement is detected by means of a sensor. This ensures that the axial z-displacement, which is included in the correction of the lateral x / y positions of the point objects, is always known exactly.

The shifting of the depth of field relative to the sample can take place in that either the sample is moved relative to the imaging optics or the imaging optics are moved relative to the sample along the optical axis. However, the invention is not limited thereto. For example, it is also possible to use a deformable lens, a deformable mirror, a spatial light modulator or the like to shift the depth of field in the object space along the optical axis of the imaging optics. In principle, it is possible to shift the depth of field in any manner.

In a particularly preferred embodiment, the z-position of the respective point object is determined along the optical axis by determining a characteristic of a spot representing the point object in the respective sample image and assigning the z-position to this characteristic on the basis of predetermined mapping information. By way of example, a variable is taken into consideration which, as in the aforementioned astigmatism method, characterizes the shape of the light spot representing the point object. Alternatively, it is also possible to use a characteristic which, as in the case of the known bi-plane method, relates the extents of two light spots which originate from the same light spot and are generated on detection spots whose associated sharpness planes are present in the object space z-direction are offset from each other.

Preferably, the z-positions of the point objects determined in the sample image are compared with the z-positions of the same point objects determined in the further sample image as a function of the predetermined axial z-displacement. As a function of this comparison, a z-correction information is then generated on the basis of which the z-positions of the point objects determined as a function of the assignment information are corrected.

This refinement avoids the problem that the assignment information determined in advance of the actual light-microscopic measurement, which makes possible an association between the parameter determined in the measurement and the axial z-position of the point object, is often so inaccurate that a precise determination of the z-position. Position is difficult. On the basis of the z-correction information, the (inaccurate) assignment information can be corrected in this case.

According to another aspect of the invention, a light microscopic device for locating point objects according to claim 14 is provided.

The invention will be explained in more detail below with reference to FIGS. Show:

1 a schematic representation of a light microscope device according to the invention;

2 two schematic cross-sectional views, one of which shows a distortion-free PSF and the other view shows a coma distorted PSF in the object space;

3 two schematic representations showing a three-dimensional, tubular structure without coma;

4 two schematic representations accordingly 3 showing the tubular structure with coma;

5 a schematic representation showing an X-shaped structure together with a depth of field and their image;

6 a schematic representation showing the X-shaped structure and its image in a first step of an exemplary method according to the invention;

7 a schematic representation showing the X-shaped structure and its image in a second step of the method;

8th a schematic representation showing the X-shaped structure and its image in a third step of the method;

9 a schematic diagram showing the X-shaped structure and its image in a fourth step of the method;

10 a schematic view showing the X-shaped structure and its image in a fifth step of the method;

11 a graph showing the lateral x-positions of a comparison structure in the in the 6 to 10 shows steps shown; and

12 a graph showing an assignment function according to the invention, the function values of which indicate lateral x-position deviations as a function of the z-position.

1 shows in a purely schematic representation of a light microscope device 10 with an imaging optics 12 that a sample 14 on a detector 16 maps. The sample 14 is on a sample carrier 28 arranged.

The imaging optics 12 has a depth of field 18 which has an axial extent t along an optical axis O. In the following it should be assumed that the optical axis O is parallel to an axis z of a Cartesian coordinate system whose further axes x and y are arranged perpendicular to the optical axis O.

The furnishing 10 further comprises a control unit 20 that the overall operation of the facility 10 controls. In particular, the control unit has 20 via computing means, which performs the necessary for localization calculations and evaluations. The control unit 80 also controls a piezo actuator 22 on, with which the imaging optics 12 along the optical axis O to the depth of field 78 along the optical axis O, ie to move in the z-direction. One with the control unit 20 coupled sensor 24 detects the z-displacement to the imaging optics 12 and thus the depth-of-field area is moved within the object space.

The sample 14 contains a variety of point objects 26 , which are formed by fluorescent markers that adhere to the structures to be imaged. During the micrograph, the point objects become 26 individually as light spots on the detector 16 displayed. The light spots thus generated become the localization of the point objects 26 in the control unit 20 evaluated.

The device 10 to 1 is used in the sample 14 contained point objects 26 in the object space 28 to locate. The control unit determines this 20 the device 10 both the z position of the respective point object along the optical axis O and the lateral x / y position in a plane perpendicular to the optical axis O plane.

In the following it will be explained how a lateral field distortion, as caused for example by a coma, is corrected according to the invention. It should first with reference to the 2 to 4 to illustrate how lateral image distortion affects accuracy with point objects 26 be localized in the lateral direction.

In 2 is illustrated how the coma leads to a distortion of the respective point object associated PSF in three-dimensional space. It is in the left part of the image 2 First, the ideal case of a faultless PSF 30 which are radially symmetrical along their in 2 with z 'designated longitudinal axis. The longitudinal axis z 'of the PSF 30 should parallel to the optical axis O and thus to the z-axis of in 2 are shown coordinate system.

The coma now leads to a distortion of the PSF 30 through which the radial symmetry of the PSF 30 is broken, as in the right part of the picture 2 is shown. This breaking of the radial symmetry leads to a lateral shift of the center of gravity of the PSF 30 in the xy plane. Again 2 can be seen directly, the lateral center of gravity shift is dependent on the z-position.

The coma leads to the fact that the centers of gravity of the PSFs assigned to the individual point objects are no longer in the same x / y position at different z positions. A corresponding effect is obtained by considering a tilting of the PSF in space instead of the coma.

If there are aberrations that result in the above-mentioned effect of the center of gravity shift, this leads to image errors in the high-resolution sample image, such as a comparison of the 3 and 4 shows. 3 shows the picture of a three-dimensional, tube-like structure 32 extending along the z-axis. The structure 32 should be colored, for example, with a suitable dye whose molecules are localized in the manner explained above. From the determined and stored position information, a high-resolution image is then assembled. This is in the right part of the picture 3 for an exemplary two-dimensional projection of all point objects located in three-dimensional space. In this two-dimensional projection, the considered structure is represented by a circle.

While 3 shows the error-free case illustrated 4 the effect of lateral field distortion. Thus, lateral field distortion results in molecular positions that are located in different z-positions but in the same x / y position being erroneously set to different x / y positions, like the left-hand image of the 4 shows. Thus, in the two - dimensional projection, a deviation from the original circular shape can be seen, as the right partial image of the 4 shows.

With reference to the 5 to 12 is explained below by way of example how the above-described lateral field distortion is corrected according to the invention.

5 shows in the left part of the figure an X-shaped structure 32 which extends by way of example in the xz plane. The X-shaped structure 32 should be colored with a suitable dye, which is the point objects to be located 26 (see. 1 ) form. To localize these dyes, their positions within the in 5 depth of field represented as a rectangular area 18 determined. The resulting, high - resolution image is in the right part of the 5 shown. Thus, the right figure part shows the image space optically conjugate to the object space shown in the left figure part.

In the following it is assumed for the sake of simplicity that the imaging optics 12 caused a field distortion in the x-direction alone. The resulting image distorted in the x-direction is shown by solid lines. In contrast, the image that would result in an ideal image without image field distortion, shown in phantom.

In 5 is also the central focus or focal plane of the depth of field 18 With 34 designated. It should be noted that in both the left and the right part of the 5 (and also in the others 5 to 10 ) are given the same reference numerals, although the left figure part represents the object space and the right figure part represents the image space optically conjugate thereto.

In the 6 to 10 is shown as the depth of field 18 in each case by a z-displacement .DELTA.z in the object space along the optical axis O is shifted in several steps to the inventive correction of the lateral x / y positions of the point objects 26 make.

That in the 6 to 10 The method illustrated is based on the following consideration. As a result of the lateral field distortion in the x-direction, point objects that are within the depth of field 18 Although in the same x-position, but lie in different z-positions, in which by the imaging optics 12 generated image displayed in different x-positions. However, it is possible to have a z position within the depth of field 18 to define, in which the x-position assignment between object space and image space is by definition correct. This z-position within the depth of field 34 thus forms a reference position. The xy plane, which is located in the object space at this z position, accordingly defines a reference plane. In the present embodiment, the central focus plane becomes 34 the depth of field 18 used as such a reference plane. The in the 6 to 10 The sequence of steps shown now has the goal of generating an assignment function as correction information which depends on the distance of a point object from the reference plane 34 provides a correction of the x-position that just compensates for the mislocalization in the x-direction caused by the lateral field distortion.

The mapping function is generated in the present embodiment by not just an image of the X-shaped structure 32 , but a total of five images, each with .DELTA.z shifted depth of field 18 be generated.

The exemplary correction procedure begins in step 1 6 in which the focus plane acting as the reference plane 34 the depth of field 18 is located within the object space along the optical axis O in the position z1. The right part of the 6 shows the high-resolution image resulting from this arrangement.

Subsequently, in step 2 7 the imaging optics 12 and thus their depth of field 18 shifted by the known z-displacement Δz (= z ₂ - z ₁ ) along the optical axis, so that the reference plane serving as a reference plane 34 enters the position z2. Again, in this arrangement, a high resolution image is created with the positions located in the z direction

are corrected by the amount Δz so as not to shift the resultant image by Δz as compared with the image formed in the previous step 1. In the resulting image is the X-shaped structure 32 is shown distorted for all z positions except z2 in the z-direction, as can be seen again by the ideal structure shown in dashed lines. How the comparison of 6 and 7 shows, the distortion in the high-resolution image changes with the shift of the depth of field 18 because the X-shaped structure is located within the depth of field 34 changes.

As in the 8th to 10 is shown, then in further steps 3, 4 and 5, the depth of field 18 successively each shifted by the predetermined z-displacement .DELTA.z along the optical axis O. That's what the reference plane feels 34 as it were the X-shaped structure 32 in the object space along the optical axis O from.

For example, after the generation of the five high-resolution images, the part of the X-shaped structure becomes 32 in step 3 in the reference plane 34 is selected as comparison structure. This comparison structure is then also identified in the other images generated in steps 1, 2, 4 and 5. Then, the lateral x-position of the comparison structure is determined based on the individual images generated in the various steps. Since the comparison structure in the recording of the individual images relative to the depth of field 18 in different z-positions, different x-positions result from image to image as a result of the image field distortion in the x-direction. These are in the graph after 11 for the individual steps 1 to 5 shown.

The desired assignment function is finally obtained by plotting the lateral x-position deviations of the comparison structure determined in steps 1, 2, 4 and 5 against the x-position of the comparison structure determined in step 3 against the associated z-positions. This is in the graph after 12 shown. The five function values of the assignment function assigned to steps 1 to 5 can then be used as interpolation points in order, for example, to generate intermediate values by way of interpolation in order finally to obtain a continuous assignment function. With such an assignment function, the lateral position deviations for all z positions can be specified to produce a distortion-free, high-resolution image.

LIST OF REFERENCE NUMBERS

10

Light microscopic device

12

imaging optics

14

sample

16

detector

18

Depth of field

20

control unit

22

adjustment

24

sensor

26

point objects

28

sample carrier

30

PSF

32

tubular structure

32

X-shaped structure

34

reference plane

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• WO 2006/127692 A2 [0002, 0004]
• DE 102006021317 B3 [0002, 0004]
• WO 2007/128434 A1 [0002]
• US 2009/0134342 A1 [0002]
• DE 102008024568 A1 [0002]
• WO 2008/091296 A2 [0002, 0004]

Cited non-patent literature

• "Sub-diffraction-limiting imaging by stochastic optical reconstruction microscopy (STORM)", Nature Methods 3, 793-796 (2006), MJ Rust, M. Bates, X. Zhuang [0002]
• "Resolution of lambda / 10 in fluorescence microscopy using fast single molecule photo-switching", Geisler C. et al, Appl. Phys. A, 88, 223-226 (2007) [0002]
• Particle-tracking experiments were known as described in Kajo et. al., 1994, Biophysical Journal, 67, Holtzer et. al., 2007, Applied Physics Letters, 90 and Toprak et al., 2007, Nano Letters, 7 (7). They have also been used in imaging techniques based on the switching and localization of single molecules described above. Reference is made to Huang et al, 2008, Science, 319 and Juette et al., 2008, Nature Methods. The prior art is further described in Pavani et al., 2009, PNAS, 106 [0009]
• "Astigmatism Methods" (the above-referenced Kajo et al., Holtzer et al., And Huang et al.), "Bi-Plane" methods (see Toprak et al., And Juette et al.), And "Double helical method" (see Pavani et al.) [0013]

Claims (15)

Light microscopic method for the localization of point objects ( 26 ) in a sample ( 14 ), in which the sample arranged in an object space ( 14 ) by means of an imaging optics ( 12 ), which in the object space has a depth of field ( 18 ) has predetermined axial z-dimension (t) along its optical axis (O), to a detector ( 16 ) is mapped; and those in the sample ( 14 ) contained point objects ( 26 ) within the depth of field ( 18 ) can be localized on the basis of a sample image obtained by imaging the sample ( 14 ) on the detector ( 16 ), lateral x / y positions of the point objects ( 26 ) are determined in the direction perpendicular to the optical axis (O); characterized in that the depth of field ( 18 ) within which the point objects ( 26 ) in the object space relative to the sample ( 14 ) along the optical axis (O) at least once by a predetermined axial z-displacement (Δz) smaller than the axial extent (t) of the depth-of-field ( 18 ) is shifted, and at axially shifted depth of field ( 18 ) the sample ( 14 ) by means of the imaging optics ( 12 ) again on the detector ( 16 ) and another sample image is generated; based on this further sample image again the lateral x / y positions of the point objects ( 26 ) be determined; lateral x / y position deviations between the lateral x / y positions of the same point objects determined on the basis of the different sample images ( 26 ) be determined; and in dependence on the determined lateral x / y position deviations a correction information is generated, on the basis of which the lateral x / y positions of the point objects ( 26 ) corrected based on at least one of the different sample images. Light microscopic method according to claim 1, characterized in that within the depth of field ( 18 ) at least one perpendicular to the optical axis (O) lying reference plane ( 34 ) defined when moving the depth of field ( 18 ) stationary relative to the depth of field ( 18 ) remains; one of the sample images is defined as the reference image, and based on the reference image, a comparison structure is defined which comprises at least one of those point objects ( 26 ) which, when the reference image is recorded in the reference plane ( 34 ) of the depth of field ( 18 ) are arranged; the comparison structure is identified in the at least one other sample image; the lateral x / y position of the comparison structure is determined in each case on the basis of the different sample images; the lateral x / y positional deviation between the lateral x / y positions of the reference structure determined on the basis of the different sample images is determined; and the correction information is generated as a function of the lateral x / y position deviation determined for the comparison structure. Light microscopic method according to claim 2, characterized in that the depth of field ( 18 ) is axially displaced in several steps, in each of these steps the determined based on the associated sample image lateral x / y positional deviation of the comparison structure with respect to the lateral x / y position of the determined based on the reference image comparison structure is determined, and as correction information an assignment function whose functional values in each case specify the lateral x / y positional deviation of the associated comparison structure determined in the respective step as a function of its axial z-position along the optical axis (O). A light microscopic method according to claim 3, characterized in that values of the assignment function which are between those obtained by the stepwise shifting of the depth of field ( 18 ) determined function values are determined by interpolation. Light microscopic method according to claim 3 or 4, characterized in that the lateral x / y positions of the point objects ( 26 ) are corrected by image processing directly in the associated sample image. Light microscopic method according to claim 3 or 4, characterized in that the lateral x / y positions of the point objects ( 26 ) are corrected in a data record obtained from the associated sample image and a corrected sample image is generated on the basis of this corrected data set. Light microscopic method according to one of claims 2 to 6, characterized in that the comparison structure in the at least one other sample image as a function of the on the detector ( 16 ) detected image brightness is identified. Light microscopic method according to one of the preceding claims, characterized in that the lateral x / y position deviations are determined by a correlation method. Optical microscopy method according to one of the preceding claims, characterized in that the sum of the individual axial z-displacement paths (Δz) substantially equal to the axial z-dimension (t) of the depth of field ( 18 ). Light microscopic method according to one of the preceding claims, characterized in that the axial z-displacement path (Δz) by means of a sensor ( 24 ) is detected. Light microscopic method according to one of the preceding claims, characterized in that the depth of field ( 18 ) in the object space relative to the sample ( 14 ) is displaced along the optical axis (O) about the axial z-displacement (Δz) by moving the sample ( 14 ) relative to the imaging optics ( 12 ) or the imaging optics ( 12 ) relative to the sample ( 14 ) is displaced along the optical axis (O). Light microscopic method according to one of the preceding claims, characterized in that the z-position of the respective point object ( 26 ) along the optical axis (O) is determined by a characteristic of the point object ( 26 ) is determined in the respective sample image representing light spot and this characteristic is assigned the z-position on the basis of a predetermined assignment information. Light microscopic method according to claim 12, characterized in that the z-positions of the point objects (z) determined in the sample image ( 26 ) with the z-positions of the same point objects determined in the further sample image as a function of the predetermined axial z-displacement (Δz) ( 26 ) are compared; and depending on this comparison, a z-correction information is generated, on the basis of which the z-positions of the point objects determined in dependence on the assignment information ( 26 ) Getting corrected. Light microscopic device ( 10 ) for the localization of point objects ( 26 ) in a sample, with imaging optics ( 12 ), which in an object space has a depth of field ( 18 ) has predetermined axial z-dimension (t) along its optical axis (O), a detector ( 16 ) to which the imaging optics ( 12 ) a sample arranged in the object space ( 14 ) maps; and a control unit ( 20 ) in the sample ( 26 ) contained point objects ( 26 ) within the depth of field ( 18 ) based on a sample image, the imaging optics ( 12 ) on the detector ( 16 ), lateral x / y positions of the point objects ( 26 ) in the direction perpendicular to the optical axis (O); characterized in that one by the control unit ( 20 ) controlled adjusting unit ( 22 ) the depth of field ( 18 ) within which the point objects ( 26 ) in the object space relative to the sample ( 14 ) along the optical axis (O) at least once by a predetermined axial z-displacement (Δz) smaller than the axial extent (t) of the depth-of-field ( 18 ), and the imaging optics ( 12 ) with axially shifted depth of field ( 18 ) the sample ( 14 ) again on the detector ( 16 ) and generates another sample image; the control unit ( 20 ) again based on the further sample image, the lateral x / y positions of the point objects ( 26 ) determined; the control unit ( 20 ) lateral x / y-position deviations between the lateral x / y-positions of the respective same point objects, determined on the basis of the different sample images ( 26 ) determined; and the control unit ( 20 ) generates correction information in dependence on the determined lateral x / y position deviations, on the basis of which the control unit ( 20 ) the lateral x / y positions of the point objects ( 26 ) corrected based on at least one of the different sample images. Light microscopic device ( 10 ) according to claim 14, characterized by a sensor ( 24 ) for detecting the axial z-displacement of the depth of field (FIG. 18 ).

Images